تعهد نامه
نوع مقاله : مقالات پژوهشی
عنوان مقاله English
نویسندگان English
Background and Purpose: This study investigates and predicts the concentration of PM10 pollutant in Mashhad using simple statistical techniques and also the LSTM model with a focus on traffic restrictions before and during the COVID-19 pandemic.
Materials and Methods: First, data related to the concentration of PM10 pollutant were collected from air pollution monitoring stations in Mashhad. Then, using a paired t-test, the statistical changes in PM10 concentration before and during the quarantine period were investigated. Also, the LSTM machine learning model was used to predict the effect of quarantine on PM10 levels during this period, which included data processing, model training, and evaluation of prediction accuracy using various criteria.
Results: The results of the paired t-test showed a 16% decrease in the average concentration of PM10 during the quarantine period, which is specified by a mean difference of 4.397 μg/m3. Although this decrease was not statistically significant, the relative improvement in air quality during this time period is remarkable. Also, in the study of the 210-day period before and after COVID-19, the results showed the significant impact of quarantine measures on air quality, and these changes did not occur randomly. In the next step, the LSTM machine learning model was used to predict the effect of quarantine on PM10 levels. The value of the coefficient of determination (0.8) indicates a strong correlation between the predictions and the actual concentration of PM10.
Conclusion: The values of mean square error (3.01) and mean absolute error (2.56) also indicate the high accuracy of the LSTM model predictions and their proximity to the actual values. These results demonstrate that the LSTM model has been able to predict the concentration of PM10 pollutant with high accuracy and confirms its high efficiency in analyzing time series data.
Open Access Policy: This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. To view a copy of this licence, visit https://creativecommons.org/licenses/by/4.0/
کلیدواژهها English